Rama de producción nacional y representatividad

As noted in section 2.5 above, in the present literature the distinction between measures of KT and measures of KT impact can be somewhat blurry. Even when studies purportedly attempt to measure the extent of KT transfer, the direct process of KT is almost never directly measured, but rather inferred on the basis of occurrence of certain outcomes such as funding, patents and licenses, which can just as easily be conceived of as outcomes/impacts, and even inputs (Caro, de Lucio & Gracia, 2003). Some of the commonly studied outcomes/impacts are presented in this section.

Virtually all measures of the impact of KT are measures of economic impact/value added, either at the firm level (e.g. new products, processes) or the social level (e.g. social returns). This is expected, considering that knowledge is a primary driver of economic growth and that this understanding lies behind all policy efforts to intensify U-I KT. However, it has also led to an under-emphasis on indirect/intermediary impacts of KT such as learning and improvements in human capital and research capacity, which are important intermediary impacts of KT at the firm level and may be part of the mechanisms for the overall economic impacts of KT.

The absence of well-developed measures of the intermediary processes of KT results in KT impact being inferred indirectly, either by collecting aggregate innovation/financial data, or by soliciting information from industrial companies about specific contributions made by KT activities on firms’ activities, products and profits. There is a considerable research tradition in the measurement of knowledge impacts of universities on firms. The oldest in this tradition are studies of “social impact” of university research, e.g. Griliches (1958) compared the cost of hybrid corn research with the value of increases in corn production that resulted from advances in hybrid corn research, and estimated a rate of return of 700%.

Rather than attempting to detect the social returns (arguably the “ultimate” measure of impact) from KT, some studies attempt to estimate more directly the role of KT on firm performance. Based on the responses to a survey of 76 US firms, Mansfield (1998) estimated that about 10% of new products and

processes would have been introduced with a great delay without contributions from academic research. Various newer studies (e.g. Beise & Stahl, 1999; Monjon & Waelbroeck, 2003) have used the same or similar measures of impact, i.e. the self-reported contribution of academic knowledge to the invention of new products or improvements in existing ones, and generally find positive relationships. Such information is gleaned through surveys, which makes it somewhat unreliable due to its dependency on identifying a specific person within an organization who can provide this information (typically a R&D manager or a business owner in the case of SMEs).

Some studies relying on existing data attempt to capture intermediary impacts of public research on private companies. Nelson (1986) uses firms’ ratings of significance of university research in general and finds that more R&D intensive firms tend to value university research higher, which Nelson interprets as evidence that one of the primary impacts of university research is to increase the productivity of private R&D thus facilitating the discovery and exploitation of opportunities for innovation (rather than generate new technology per se), a point further explored elsewhere (Pavitt, 1998; Rosenberg & Nelson, 1994).

An extension of this line of reasoning and a modified dependent variable is proposed by Kaufmann and Tödtling (2001) who show that firms who interact with universities improve their ability to innovate, defined as introduction of products that are new for the firm; in addition, the effect was stronger for radical rather than incremental innovations. A variety of KT impact variables is proposed by Adams and colleagues (2003) in a study of the effects of Cooperative Research and Development Agreements between firms and federal laboratories in the United States. The variables include: in-house company financed R&D budget (i.e. in-house R&D spending) and number of patents.

An indirect approach to study the impact of public science and particularly the impact of university- industry collaborations is to study the characteristics of award winning inventions (Block & Miller, 2008). Specifically, Block and Miller discover that the proportion of award-winning innovations that involve some form of public-private collaboration has drastically increased since the 1970s.

One of the recent attempts to measure the impact of U-I KT (Arvanitis, et al., 2008) aims to differentiate the impacts that different KT channels may have on firm performance. The dependent/impact variables are: i) sales of new products as a percentage of total sales; and ii) sales of significantly improved existing products as a percentage of total sales, modelled as a function of the full spectrum of KT activities that firms may engage in. The results suggest that firms with primarily research-oriented collaboration with universities perform better on those variables.

Other approaches (Cohen, et al., 2002; Schartinger, et al., 2002), shift the focus from products in general, and instead ask companies to rank the importance of different stakeholders (e.g. suppliers, customers, public research institutes) or types of KT channels for their innovation activities. Cohen and colleagues (2002) differentiate between the contribution of public research to new project ideas and contributions to solving existing problems, and find that public research is more useful in the latter than the former. That finding is qualified by a separate study (Meyer-Krahmer & Schmoch, 1998), which suggests a moderating effect of sectoral differences: in science-intensive fields, private companies primarily value public research as a general knowledge source, while in less science-intensive industry, the primary value of public research is in problem solving.

Overall, the generic types of impacts of KT can be grouped in four broad categories: organizational outcomes, knowledge outcomes, technological outcomes and financial outcomes (Tsai-Lung, 2001). Such distinctions should be kept in mind when designing measures of KT, insofar as these categories can be confused with each other: for example, while the improvement of a firm’s knowledge base may (and ideally should) improve its competitive position, financial or technological improvements may be an indirect effect of a KT interaction specifically aimed at augmenting a firm’s knowledge base (which is one of the leading motivations for firms to enter U-I KT). Becker (2003) applies a useful and simpler distinction: universities can influence either or both:

• Firm-level innovation inputs (e.g. learning, absorptive capacity, internal R&D investments) • Firm-level innovation outputs (e.g. products, sales, patents, etc.)

This distinction will be useful to consider when analyzing different KT channels, and indeed, whether a KT channel targets a firm’s innovation outputs or inputs is conceptually useful as one of the major dimensions characterizing KT channels.

Similarly, the issue of the role of HR management practices in KT (see next section) has been much more intensively researched in the firm/private sector context than in the U-I context (for a recent overview, see Smale, 2008). Many of the business literature lessons may be applicable in the university- industry context, although it should be emphasized, again, that unlike private companies who have direct authority over their personnel, U-I KT remains a discretionary behavior and thus not all traditional HR management practices are directly applicable.